My first twenty-games-in article was kind of a bummer. Objectively, the Caps aren’t playing great hockey right now. There’s really no way to get around that fact, but there’s also no reason to get hung up on it either.

I have a suspicion that fans and analysts tend to overestimate how far a bad team is from becoming a good team. My pet theory is that any bad team isn’t more than three years from contending for a championship – if they make smart decisions. Making smart decisions starts with identifying problems, and today’s exercise is good at identifying problems.

Let’s look at how each forward plays with each defenseman, and maybe we can figure it out.

This a visualization invented by Tyler Dellow, now of The Athletic. This is our fourth year doing the vis, and I think it’ll be even more useful now than it has been in the past.

Quick overview:

• Each defenseman is a column and each forward is a row.
• Where the row and column intersects, the number means what percentage of the total 5-on-5 shot attempts belong to the Caps when those two players are on the ice.
• The cells are color-coded: red is bad, green is good, white is even.
• The rows and columns are sorted by ice time, so the top-left players are on the ice a lot, and the bottom-right players are on the ice less.
• I’ve also added a second table below that shows the number of 5-on-5 minutes that the players share. It’s also color-coded so that it gets yellow the less that forward plays with that defenseman. If the color is purple, the sample size is very small (e.g. due to a player injury) and therefore the shot-attempt is less reliable.

Okay, here we go.

Shot Attempt Percentage

Time on Ice

Observations

• What we want to see in the first table is bright green in the top left, which would mean our most used players getting the most offense. We see that a little, but it’s interrupted by lots of patches of deep reds (i.e. at or below 40 percent of shot attempts), and those are worrying.
• The deepest red is Brooks Orpik, whose play is the single biggest problem for the 2017-18 Washington Capitals. Whether by his increased ice time (up 4.5 minutes per game from last year!), tougher opponents (often the top line), or advanced age (he turned 37 two months ago), Orpik is getting consistently crushed during 5-on-5. He’s not just under 50 percent; he’s mostly under 40 percent – the Mendoza line. For context, no defenseman who played over 300 minutes last season was under 40, though Anaheim’s Korbinian Holzer and Buffalo’s Rasmus Ristolainen came closest. This is not good company.
• Nearly all top-six forwards see their worst 5-on-5 performance when playing with Orpik, but apart from him the numbers get greener – mostly. Let’s investigate that.
• John Carlson has spent two-thirds of his 5-on-5 minutes next to Orpik, and it’s been a bloodbath. When Carlson is not with Orpik, his 42.5 percent SA% jumps up to 61.5 percent – even with Carlson playing tough minutes and lots of them. So Carlson’s column is iffy, but there are reasons why.
• Matt Niskanen‘s numbers are a mixed bag, mostly due to the 13 games he missed due to injury. He’s spent less than a half hour with all but two forwards, Oshie and Backstrom, so I think it’s best to defer judgment for now.
• Evgeny Kuznetsov had some of the best possession numbers on the Capitals last season, but now he’s struggling with pretty much all defensemen. His most frequent linemates offer a suggestion why: last year they were Marcus Johansson and Justin Williams; this year they’re Alex Ovechkin and DSP.
• Devante Smith-Pelly should not be a top-six forward, but he really hasn’t been bad except when out with Orpik. Using Natural Stat Trick’s line tool, we can see that the Ovi-Kuzy-DSP line gets 53.3 percent of the shot attempts in the hour they’ve been free of Orpik. The problem is that they get used with him so much: 45 percent of the time.
• And perhaps that speaks to underlying reason the Caps are relying to much on Orpik to their own detriment: the defensive depth is atrocious. Taylor Chorney is in the bottom 10 percent of NHL defensemen in shot-attempt percentage and expected-goals percentage. His play with every forward except Oshie and Backstrom is not just bad, it’s in the league basement. He’s got the most talented forward on the planet getting outplayed 60-to-40. The third line is barely playing a one-third of their time on offense.
• Aaron Ness doesn’t have enough ice time to merit inclusion in the tables above, but his results line up with Chorney’s: almost unspeakably bad aside from time with Backstrom and Oshie.
• And that’s a pattern by the way. While Nicklas Backstrom and TJ Oshie have had a hard time producing goals during 5-on-5, they’ve so far resisted the downward drag of the team’s trajectory in shot attempts. Defensemen typically see their best 5-on-5 numbers when they’re lucky enough to take a shift with the Backstrom line, which means Orpik sees a tolerable 49.0 percent SA% and Djoos sees a dominant 58.4 percent. If you’re looking for bacon bits (i.e. they make everything better), it’s here.
• Madison Bowey‘s column is spotty, and I think that speaks to his usage and inexperience. I’ve seen solid decision-making and gap control from him, especially on neutral ice, but I don’t think that either Dmitry Orlov or Brooks Orpik are a particularly good complement for him.
• Neither to my eyes or in the numbers does Christian Djoos look like a rookie.
• Jay Beagle has a very tough job, and we shouldn’t expect him to pull the team’s defensive numbers up considering his role (lots of defensive-zone starts). Still, considering his own precedent, he is having a down year.

That’s a lot of information, but put together I think it all informs one pattern: while the core of the team (Backstrom, Ovi, Carlson, Orlov) still seems good, the Capitals have profound depth problems on their roster, particularly on defense. Those problems could be partially mitigated with changes in deployment and usage (i.e. coaching), but they cannot be wholly solved without personnel changes (i.e. management).

For the moment, Barry Trotz and Brian MacLellan are in detente. Barry Trotz is not making any substantial changes to his lineup despite overwhelming evidence that it’s not working, and Brian MacLellan has appealed to the Washington Post’s most credulous columnist to help him manage expectations. Until one of them blinks, there will be no improvement.

Headline photo: Amanda Bowen